Across our projects, we hear a clear shift in the conversations around digitalisation, from GenAI as a creator of knowledge to GenAI as a driver of action. Discussions about data capture and infrastructure that were lukewarm a decade ago are now heating up. Across industries, experts are asking a simple question: Do we actually have the data foundation to build on?
Yet when we speak with people close to the ground, building, scaling and using, organizational data systems, a different sentiment keeps coming up. Beneath the excitement around AI lies a growing fatigue: Not with the technology itself, but with the drowning feeling of endless data, full visibility and dashboards overflowing with information. As one participant described the challenge of digitalising supply chain management:
“Honestly.. I don't need more data and visibility, I need someone to provide me visibility and then the ability to take action on that visibility. Visibility is great, but it does nothing for you if you're not able to take action on it.”
Once hailed as the new gold, data has become so abundant it’s overwhelming. With dashboards multiplying by the day, the paradox is clear: more data hasn’t brought more clarity - if anything, it’s clouded it.
Chasing Data, Missing Clarity
“…most businesses are already data rich, but insight poor.”
— Bernard Marr, 2015
Today’s organizations are obsessed with “data plays”: collecting, refining, and monetizing information in the pursuit of growth and control. Raw data has become a kind of status symbol, a mark of sophistication and power. Scholars even have a term for this: data fetishism.
At the same time, we as consumers have grown used to trading our personal data for convenience. We tracked our steps and sleep, previewed furniture in our homes in AR, and shared our spending patterns with budgeting apps. Today, we can trade data for convenience through voice assistants in our homes or cars, or even share our biometric data with “try-on” apps that recommend skincare. Data flows freely through our daily lives, offering a sense of empowerment.
Or perhaps just the illusion of it.
In practice, many of these data-driven promises fall short. What we often get is visibility, not clarity. Parcels can be traced but not foreseen; fertility trackers offer probabilities, not certainty; and banking apps flood us with notifications that inform, yet rarely enlighten or help us act.
The real challenge we hear organizations and experts face today isn’t about gathering more data. It’s about moving from visibility to actionability. About turning what’s already there into meaningful insight. Having precision with purpose.
It’s not “Bad Insight”, it’s Bad Logic.
When we speak with teams struggling to act, they often blame poor insight. The insight they get is partial, low quality, unvalidated, untimely or just somehow intuitively wrong. And the teams responsible for generating those insights usually point to bad data, i.e. data that’s inaccurate, incomplete, or irrelevant.
But the real issue? That often lies elsewhere.
Action is only as effective as the insight that informs it, and insight only as sound as the logic that supports it. The problem isn’t the insight or even the data, it’s the logic supposed to connect the two. That is, the interpretation and shaping of the data with purpose, context, and relevance to form relevant insight.
One retail leader we spoke with illustrated this perfectly:
“We asked developers from Silicon Valley to build us a custom AI model for our retail operations. But no matter how much we tried, they couldn’t get it right. The model kept over-estimating and was completely unusable in action — they simply didn’t grasp our industry specifics; purchase patterns and changing trends.”
As the retail example shows, developing clear logic starts with understanding the specifics that make an insight work in its real-world context. Logics are not universal; they are particular. They are shaped by where and how it’s applied. Industry, geography, target users, and organizational goals all influence what counts as logical for a particular situation or system.
And to build robust logic, you must understand what would make an effective insight.
In our work, we often describe insight through four key qualities, each designed to guide rather than merely inform:
- Insight prompts meaningful action — by reframing the problem to allow new solutions.
- Insight arrives when it matters — by anticipating what will be needed next.
- Insight resonates with experience — by turning the strange into the familiar.
- Insight is specific — by reflecting the unique dynamics of a given context.
For example, the insight of a parcel update is only useful when it fits seamlessly into someone’s day. When it helps them plan or adjust their schedule without friction. Simply knowing that a parcel is “on the way” adds little value; what matters is when that information arrives, how it aligns with the user’s needs, and what decisions it enables them to make.
Similarly, a menstrual tracker becomes truly insightful when it doesn’t just record what has already happened, but when it flags gaps, indicates cycle changes, or provides personalised recommendations along the next cycle. In doing so, it shifts from being a passive recorder of information to an active guide that supports timely, meaningful action.
Building true insight isn’t just about putting data together. It’s about applying a clear, context-led logic, a method, to gather and connect information to make sense to the user. One that helps us uncover what we need to know while understanding clearly how we know it. When the created logic is contextually congruent, insight becomes actionable.
In short, it’s not more data that drives insight and action. It’s logic, or rather, logics.
Time to Build Your Logics
True clarity comes from the specificity of applied logic: understanding the context, the purpose, and the patterns that matter. When logic is tailored to purpose, insight becomes powerful. That’s why the most significant AI innovations emerge within verticals: algorithms built for accounting, supply chain forecasting, or biotechnology succeed because they embed the nuances and rhythms of their domain.
What if your organisation shifted their focus from collecting data to developing your epistemologies – the particular logics, heuristics, and frameworks of their expertise area to guide and explain how knowing is produced and validated.
Interesting questions to ask could be simple yet fundamental ones:
- What counts as knowledge in our context?
- Which contextual factors should shape the logic we build?
- How do we know we know, and not merely believe?
If you’re interested in exploring questions like these, to ensure your data builds sound insights, which in turn direct meaningful action, we would love to share more inspiration. Reach out to Marianna and Carolina.
.png)
